Fw

CC

a

AA

KFPHW

1

tschocwmtiigb

mpattt

0d

Catalysis Today 157 (2010) 119–124

Contents lists available at ScienceDirect

Catalysis Today

journa l homepage: www.e lsev ier .com/ locate /ca t tod

e-modified TiO2 photocatalysts for the oxidative degradation of recalcitrantater contaminants

.A. Castro-López, A. Centeno, S.A. Giraldo ∗

entro de Investigaciones en Catálisis (CICAT), Escuela de Ingeniería Química, Universidad Industrial de Santander (UIS), Cra. 27 Calle 9, Bucaramanga, Colombia

r t i c l e i n f o

rticle history:vailable online 7 August 2010

eywords:e–TiO2

hotocatalytic oxidationydrothermal synthesis

a b s t r a c t

Fe–TiO2 with enhanced photocatalytic oxidation activity was synthesized by hydrolysis of titaniumbutoxide. Sol–gel process leads to less active Fe–TiO2 particles than bare TiO2; meanwhile, the hydrother-mal synthesis of Fe–TiO2 seems to promote the stabilization of Fe3+ in the TiO2 structure promoting thephoto-oxidation activity of TiO2 towards the degradation of azo-dyes and the inactivation of Escherichiacoli (E. coli) in water. Photocatalysts were characterized by X-ray diffraction (XRD), diffusive reflectancespectrophotometry (DRS), and N2 adsorption–desorption. The TiO2 synthesized by the hydrothermal

ater contaminantssynthesis reaches the TiO2 Degussa P-25 photo-oxidation activity towards the degradation of orange II.Furthermore, Fe loading of TiO2 in hydrothermal synthesis improves remarkably the photoactivity of HT.The comparative performance of the photocatalysts towards the direct photo-oxidation of iodide, thephotocatalytic degradation of orange II and the photocatalytic disinfection of water infected with E. colisuggested different mechanisms of oxidation: a surface-hole-oxidation mode for hydrothermally syn-

d a h •

thesized TiO2 particles anin the TiO2 P25 surface.

. Introduction

The photocatalytic oxidation is a recently proposed route forhe production of drinking water, especially in zones with higholar irradiations and lack of water supplies. In photocatalysis, theapacity of a UV-A-activated photocatalyst, such as TiO2, to produceighly oxidant radicals is used to efficiently oxidize a wide varietyf organic molecules [1,2] and inactivate bacteria [3,4]. These radi-als are produced by redox reactions of water and dissolved oxygenith the UV-A photogenerated charged species of TiO2: the pro-oted conduction band electron (e−) and its positive counterpart,

he valence band hole (h+). Among the oxidative radicals producedn TiO2 illumination, the hydroxyl radical (•OH) is responsible,n most of the cases, for the complete mineralization of the tar-et molecule [5], as well as the complete inactivation of differentacteria [3].

However, the efficiency of TiO2 is decreased when the UV-A pro-oted charged species recombinate, and, thus, the possibility to

roduce •OH is diminished. Transition metals have been proposed

s dopants capable of enhancing the photoactivity of the TiO2 byrapping the e−–h+ pair [6,7]. In particular, iron doping increaseshe •OH production by its capacity to act as h+ traps, which reduceshe recombination [7–12], and, thus, the chance for the e−–h+ to

∗ Corresponding author. Tel.: +57 7 6344746; fax: +57 7 6459647.E-mail address: sgiraldo@uis.edu.co (S.A. Giraldo).

920-5861/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.cattod.2010.04.050

omogeneous phase oxidation mode promoted by OH radicals produced

© 2010 Elsevier B.V. All rights reserved.

form oxidative radicals is improved. In order to modify the TiO2with Fe and achieve an increase in the photocatalytic activity,some synthesis methods in liquid phase have been used, such assol–gel [9,12,13] and hydrothermal synthesis [10,11,14]. Neverthe-less, when Fe is oxidized to Fe2O3 and embedded in the TiO2 matrix,its photoactivity decreases because this oxide acts as a recombina-tion center of the photogenerated charges [9], and thus, reducingphotoactivity. Moreover, new insights recently found propose thatinteractions between Ln2+ doped TiO2 or Fe doped TiO2 and thetarget molecule may increase the photo-oxidation, such as, thephotosensitization of TiO2 by organic dye molecules [15], and thephoto-leaching of Fe from the TiO2 matrix [16]. Both processes arenot related to the mechanism of recombination but increase thephoto-oxidation of the target.

In addition, the coexistence of phases in TiO2 has been suggestedto reduce the recombination of the e−–h+ pair [17,18]. Anatasephase, with a band gap (Eg) of 3.2 eV, is known as the most pho-toactive phase in TiO2. However, an anatase to rutile phase (Eg of3.0 eV) ratio of 3:1 shows the best photocatalytic activity of thecommercial TiO2 Degussa P-25 [17]. Such a difference in the Eg ofthe anatase and rutile phases serves as an energy gradient, whichconducts the e− from rutile to anatase, thus, avoiding the recom-

bination. Similarly, a mixed phase TiO2, constituted of anatase andbrookite phases, has a higher performance towards the degradationof 2-chlorophenol than pure anatase TiO2 [18]. Then, an increase inphotoactivity seems to be correlated to the coexistence of phasesanatase and rutile [17] or brookite in TiO2 [19].

1 alysis

tt

phppmbam1otg

2

2

lbAaaaawwfwaaTa

TaIwtIrtww7Tw

tsEp

2

tta

tr

20 C.A. Castro-López et al. / Cat

Therefore, the photocatalytic activity of the Fe doped TiO2 pho-ocatalyst is strongly dependent on the crystalline structure andhe state of Fe in the oxide matrix.

In this work, TiO2 photocatalysts modified with Fe3+ wererepared using two different synthesis methods, sol–gel andydrothermal synthesis using titanium butoxide. The physicalroperties and the photo-oxidation ability of the Fe doped TiO2hotocatalysts are correlated to elucidate the state of Fe in theatrix. The photocatalytic production of •OH radicals was followed

y means of the photocatalytic degradation of orange II (Or-II),n organic azo-dye recalcitrant to conventional degradation treat-ents [20] and the deactivation of the Escherichia coli strain ATCC

1229 used for disinfectant quality tests [21]. Additionally, thexidative potential of the photogenerated holes was analyzedowards the direct oxidation of iodide to iodine by the photo-enerated h+ [22].

. Experimental

.1. Synthesis of Fe–TiO2 photocatalysts

Two methods were chosen to synthesize Fe–TiO2 photocata-ysts: sol–gel and hydrothermal synthesis. Fe-SG was synthesizedy sol–gel synthesis as follows: titanium butoxide (Ti(O-But)4; 97%ldrich) was added dropwise to isopropanol (Isop-OH; Merck),s co-solvent, in a molar ratio Isop-OH/Ti(O-But)4 = 55. Then, andequate amount of Fe(NO3)3 was diluted in water and pH wasdjusted to 1.5 with HNO3 (65%, Merck). Fe aqueous solution wasdded dropwise to the Ti(O-But)4-Isop-OH solution. The amount ofater corresponds to a molar ratio H2O/Ti(O-But)4 = 1.5. Fe contentas adjusted to a nominal molar percentage of 0.1 (mol%). The

ormed solution was stirred until gel was formed. Then, the stirringas stopped, and the gel was aged for 72 h, and immediately dried

t 70 ◦C for 4 h. The obtained crystals were grounded in a mortar,nd the yellowish powders were calcined at 400 ◦C for 2 h. BareiO2 sol–gel synthesized, labeled as SG, was obtained as describedbove without the addition of the Fe salt.

Fe-HT was hydrothermally synthesized by hydrolysis ofi(O-But)4 in aqueous media. In this case, Ti(O-But)4 wasdded dropwise to Isop-OH, as co-solvent, in a volume ratiosop-OH/Ti(O-But)4 = 10. Then, an adequate amount of Fe(NO3)3

as diluted in water and pH was adjusted to 1.5 with HNO3. Then,he Fe aqueous solution was added dropwise to the Ti(O-But)4-sop-OH solution. The amount of water corresponds to a volumeatio H2O/Ti(O-But)4 = 25. The Fe content was nominally adjustedo 0.1 mol%. The latter suspension was hydrothermally treated withater stream in an autoclave at 120 ◦C and 1980 kPa for 3 h. Then,ater was extracted by evaporation and continuous stirring at

0 ◦C. Finally, the obtained powders were grounded in a mortar.iO2 obtained by hydrothermal synthesis, as described above butith no Fe salt addition, was labeled as HT.

All photocatalysts were named as Fe-Y, where Y stands forhe synthesis route of the Fe–TiO2 photocatalyst, SG for sol–gelynthesis, and HT for hydrothermal synthesis. TiO2 P25 fromvonik, previously known as Degussa, was used for comparisonurposes.

.2. Photocatalyst characterization

To analyze the phase structure of the photocatalysts, DRX pat-erns were collected using a D-Max IIIB Rigaku system at room

emperature from angles 2� of 2–70◦. The diffractometer was oper-ted at 40 kV and 80 mA with a monochromatic Cu K� radiation.

Specific surface areas of photocatalyst were determined byhe BET method. The N2 adsorption–desorption isotherms wereecorded on a Nova 1300 apparatus of Quantachrome.

Today 157 (2010) 119–124

UV–vis diffusive reflectance spectra (DRS) were recorded on aPerkinElmer 2034RD lambda 35 UV with an integrating sphere P/NC6951014 using BaSO4 as a blank reference. The energy band gapwidths of the samples were determined using the Kubelka–Munkphenomenological theory [23].

2.3. Photocatalytic tests

Photocatalytic activity tests were performed in 50 mL boro-silicate glass reactors illuminated in an ATLAS Suntest CPS+ systemwith external magnetic stirring. Suntest Xenon lamp’s spectrumcovers wavelengths from 300 to 800 nm, with 7% of the photonsin the UV-A region. Irradiation and temperature were controlled at400 W/m2 and 35 ◦C respectively.

2.3.1. Photocatalytic degradation of orange IIThe photocatalytic degradation of Or-II was carried out with

a 20 ppm Or-II aqueous solution with 0.25 g/L of photocatalyst.This suspension was stirred for 1 h under darkness prior to illu-mination. Samples were taken every 15 min for 1 h and centrifugedat 3000 rpm for 15 min. The concentration of Or-II in the super-natant was determined with a HP 8453 UV–vis spectrophotometerat maximum wavelength absorption, 486 nm. Blank experimentswere carried out to determine possible degradation by photoly-sis. The adsorption equilibrium concentration, reached after 1 hof stirring under darkness, was used as the initial value for thephotodegradation efficiency calculations. To follow the degrada-tion the obtained Or-II concentration data are presented as arelative concentration during testing time: C/C0, where C standsfor the concentration at any given time and C0 is the initialconcentration.

2.3.2. Photocatalytic disinfection of waterE. coli ATCC 11229 was incubated for 10 h in 100 mL of Luria

Bertani growth media (1 wt% , Tryptone from Oxoid), 0.5 wt% yeastextract from Oxoid, and 1 wt% NaCl from Merck) under 120 rpmof stirring at 35 ± 2 ◦C. A 2 mL aliquot was taken from the growthmedia for centrifugation at 3000 rpm for 15 min. The supernatantwas discarded, and cell biomass pellet was washed twice withsaline solution (0.85% NaCl); then, the pellet was suspended in steri-lized distilled water and added to a 50 mL borosilicate glass reactorfor photocatalytic disinfection. The reaction medium consists of50 mL of an E. coli cell suspension with 0.25 g/L of photocatalyst.Samples were taken from the reaction medium every 15 min for1 h and serially diluted in saline solution. To follow the concen-tration of survival E. coli 10 �L of the dilutions was plated ontoplate count agar (Merck) and incubated for 24 h before growthcolony counting. E. coli concentration data are presented as theaverage of 3 countings on plates as colony-forming units per mL(CFU/mL) from independent experiments. Initial concentration ofE. coli was adjusted to ∼1 × 107 CFU/mL. Blank experiments werecarried out to determine any possible bactericidal activity underdarkness labeled as dark, and the influence of simulated solar lightto cause solar disinfection labeled as solar disinfection (SODIS).

2.3.3. Photocatalytic oxidation of iodideThe direct photo-oxidation of iodide (I−) by TiO2 photo-

generated hole (h+) to produce iodine (I2), Eq. (1) [22], wasconducted using a 4.5 × 10−3 M solution of KI with 0.25 g/L ofphotocatalyst. Samples were taken every 20 min for 1 h and cen-trifuged at 3000 rpm.

2I− + h+ → I2 (1)

The concentration of iodine in the supernatant was followed bycomplexation of I2 with a 50 g/L starch suspension and determinedby absorption at 620 nm in a UV–vis spectrophotometer.

alysis Today 157 (2010) 119–124 121

3

3

apslTnm[ttucaotFint

FA

C.A. Castro-López et al. / Cat

. Results and discussion

.1. Characterization

The XRD patterns of the TiO2 and the Fe-modified TiO2 particlesre shown in Fig. 1. It can be seen that XRD patterns of all sam-les show an anatase-characteristic peak at 2� = 25.2◦. No corre-pondent peaks to Fe phases were identified, possibly due to theow Fe concentrations in TiO2 matrix. The ionic radius of Fe3+ andi4+ is very similar, 0.79 and 0.75 a respectively for a coordinationumber of 6, thus, iron ions may substitute titanium on the TiO2atrix or be located interstitially forming a Fe–TiO2 solid solution

14]. In hydrothermally synthesized photocatalysts HT and Fe-HT,he main peak of anatase at 2� = 25.2◦ overlaps with the diffrac-ion peaks of brookite phase at 2� = 25.34◦ and 25.68◦, which issually found when both phases coexist in TiO2 structure [18]. Aharacteristic peak of brookite is observed with a small diffractiont 2� = 30.2◦ marked as B in Fig. 1. Fe-HT shows a slight distortionf the diffraction peaks indicating a reduction of crystallinity dueo the presence of Fe. This result is in accordance with the fact that

e3+ may be inserted in the TiO2 matrix as a substituent for Ti4+ ornterstitially located. In SG and Fe-SG cases, the anatase phase isotorious for both samples with a high crystallinity. This high crys-allinity of anatase phase is related to a high photocatalytic activity

ig. 1. XRD patterns of TiO2 (HT, SG, P25) and Fe–TiO2 (Fe-HT, Fe-SG) samples. A:natase, B: Brookite, R: Rutile.

Fig. 2. DRS spectra and their corresponding calculated Eg for the TiO2 (P25, SG, HT)and Fe–TiO2 (Fe-SG, Fe-HT) samples.

of the TiO2 [24]. A slight decrease in this crystallinity is observedwhen modifying SG with Fe, suggesting an insertion of Fe into theTiO2 framework.

The DRS spectra of the TiO2 and Fe–TiO2 samples and theircorrespondent energy band gaps (Eg) are shown in Fig. 2. P25shows the highest absorption in the UV region of the spectrum(200–360 nm), while SG and Fe-SG present the lowest. When com-paring the spectrum of bare TiO2 samples, SG and HT to Fe-modifiedTiO2, Fe-SG and Fe-HT, a red-shift is notorious in the absorptionto the visible region (>400 nm). Fig. 2 shows the Eg of the photo-catalysts obtained by the Kubelka–Munk theory. The reduction inthe band gap due to the presence of Fe may be explained sinceFe induces new intra band gap states, giving to the TiO2 the capa-city to absorb light at lower energy levels [6], thus, promoting theabsorption on the visible part of the spectrum. The excitation of3d electrons of the Fe3+ transferring from the energy level of thedopant to the conduction band of the TiO2 could effectively shiftthe absorption threshold into the visible region.

The textural properties, the specific surface area (SBET), andaverage pore diameter (DA) of photocatalysts are shown inTable 1. From BET analysis, a decrease is seen in surface areawhen modifying the TiO2 with Fe loading in sol–gel (Fe-SG) andhydrothermal synthesis (Fe-HT). Hydrothermal synthesis leads toTiO2 particles (HT) with higher surface and external area andsimilar average pore diameter than SG and P25. A decrease in BETarea was found when loading TiO2 with Fe in Fe-HT and Fe-SGcases. Nevertheless, it is important to note that in photocatalysisa higher external area exposed for light absorption is preferred to ahigher surface area, since the effective area for photon absorption

is the external area of the semiconductor. Thus, for pores hidden tophoton absorption no generation of oxidative radicals is expectedinside since outer parts of the photocatalyst may screen and scatterlight just as when exceeding the concentration of photocatalyst in

Table 1Textural properties of TiO2 and Fe-modified TiO2 samples.

Photocatalyst HT Fe-HT SG Fe-SG P25

BET (m2/g) 161 147 44 41 48DA (Å) 17 13 19 N.D. 26

SBET: specific surface area, DA: average pore diameter, N.D.: not determined.

122 C.A. Castro-López et al. / Catalysis Today 157 (2010) 119–124

FPi

wo

3

Tilpmcsdttltasrbtdebtota

laheHamtpibtadftc

ig. 3. Or-II degradation as relative concentration C/C0 by: (�) Fe-SG; (�) SG; (©)25; (�) HT, and (�) Fe-HT, during 60 min of 400 W/m2 of simulated solar lightrradiation and (×) blank photolysis.

hich light penetration is obstructed [25]. But of course, adsorptionf the target molecule may be increased.

.2. Photocatalytic activity

Fig. 3 shows photocatalytic degradation of Or-II by theiO2 and Fe–TiO2 photocatalysts under simulated solar lightrradiation. Sol–gel synthesized samples, SG, and Fe-SG show theowest performance in the photodegradation of Or-II. This poorhotoactivity of sol–gel synthesized samples was not expected byeans of crystallinity of the anatase phase since SG has high phase

rystallinity. On the other hand, P25, with the lowest crystallinity,hows a high performance. Then, it is not possible to establish airect relationship between crystallinity of phases and the pho-oactivity of TiO2 without comparing the whole set of properties ofhe samples. In addition, sol–gel synthesized samples showed theowest absorption capacity in the UV region, which may explainheir poor photoactivity (Fig. 2). P25 and HT photocatalysts presentlmost the same photoactivity. Crystalline structures of each onehow two phases. In P25, anatase and rutile are present in a massatio of 3–1, and in HT, there exists a combination of anatase androokite in an undetermined mass ratio. The well-known photoac-ivity of P25 is ascribed to the coexistence of anatase and rutile [17]ue to the difference in the band gap of these phases, which is annergy gradient that forces the rutile photogenerated e− to an intraand gap energy state of anatase, avoiding its recombination withhe h+ [17]. Therefore, it is possible to assume that the photoactivityf HT may be determined by the synergistic action of the conjunc-ion of phases, anatase and brookite, as suggested by Ardizzone etl. [18].

However, light absorption capacity of HT in the UV region isower than that of P25 (Fig. 2) suggesting a decrease in photo-ctivity for HT. This, in fact, seems a contradiction since P25 and HTave similar Or-II phototodegradation activities. The textural prop-rties of the samples may add new clues for comparison purposes.T particles show a high photoactivity with the highest surfacerea. Thus, it is expected that HT particles with higher surface areaay have even more exposed area for light absorption. In this case,

here should be a high possibility for more photons to produce morehotogenerated charged species due to a higher surface extension

n HT particles, and hence, the reduction in photoactivity causedy the lower light absorption of HT may be compensated and reachhe activity of P25. On the other hand, Fe alters the photocatalytic

ctivity of the TiO2. The concentration of Fe here used is in accor-ance with the optimal concentrations of Fe in the TiO2 matrixound by Ghorai et al. [8] and Tong et al. [26]. However, Fe reduceshe photoactivity of SG despite the increase in absorption capa-ity in the visible region as seen in DRS analyses (Fig. 2). Con-

Fig. 4. Concentration of iodine achieved by: (�) Fe-HT, (�) HT, (©) P25, (�) SG and(�) Fe-SG during 40 min of 400 W/m2 of simulated solar light irradiation.

trarily, there is a remarkable increase in photoactivity in HT dueto the presence of Fe in Fe-HT, which indeed promotes visiblelight absorption. Thus, it is not possible to correlate the photo-activity of Fe-modified TiO2 particles with the increased visiblelight absorption capacity. The direct photo-oxidation of iodine maygive an explanation of the state of Fe in the TiO2 network for thephotocatalysts. The concentration of iodine reached by the TiO2and Fe-modified TiO2 photocatalysts is shown in Fig. 4. The I2concentration reached by the SG sample is slightly higher thanits correspondent Fe-modified TiO2 (Fe-SG), suggesting that Feis acting as a recombination center for the e−–h+ pair. For theFe-HT sample there is not a significant increase of the perfor-mance of HT to suggest the action of Fe as a trap for the h+.Such trap could avoid the recombination of the photogeneratedcharges and increase the direct oxidation of I− to I2 by the h+

recombination [22].Then, it is possible to assume that the chosen sol–gel route

promotes Fe as a recombination center for the photogeneratedcharges in Fe-SG. Meanwhile, the hydrothermal treatment seemsto stabilize Fe3+ species on the TiO2 matrix enhancing photo-activity. Fe doping induces some extent deformation on the crystallattice of the TiO2 due to the different atomic sizes [26]. However,hydrothermal treatment restrains grain growth, stabilizing particleformation [26], and possibly conducts iron to substitute Ti on theTiO2 matrix, thus promoting the photoactivity [11].

Adsorption of Or-II molecules on P25 and SG surfaces in the darkagitation period prior to illumination is shorter than adsorptionon HT surface indicating a strong interaction of this photocatalystsurface and Or-II. Or-II has a sulfonate functional group (–SO3

−),negatively charged, that promotes its adsorption on positivesurfaces [27]. This possibly suggests a direct oxidation of Or-II inthe surface of HT and Fe-HT by photogenerated holes; meanwhileP25, with lower Or-II adsorption, may degrade the target moleculeby oxidation due to the •OH free radicals in homogeneous phase.This assumption is in agreement with the photo-oxidation of theI− by the photocatalysts, since HT causes a higher direct oxidationof iodide than P25. Kormali et al. [28] have proposed that oxidation

activity of TiO2 is mainly due to the action of •OH radicals and to alesser extent to holes. However, the nature of the photo-oxidationmode depends on the nature of the substrate [28]. Additionally, asit is known the e−–h+ pair transfer to the media takes 100 ns–1 ms,

C.A. Castro-López et al. / Catalysis

Flt

abocsts

dopmmind4toHlHNEtepapceb

sbpdPorodaOi•

ig. 5. Survival of E. coli ATCC 11229 after 40 min of 400 W/m2 of simulated solaright irradiation using TiO2 (SG, P25, HT) and Fe-modified TiO2 (Fe-SG, Fe-HT) pho-ocatalysts.

nd the recombination takes 10–100 ns [2]. Then, it is more proba-le that surface oxidation by the photogenerated h+ on the surfacef HT takes place faster than the homogeneous phase oxidationaused by •OH homogenous phase free radicals produced on theurface of P25. Indeed, homogeneous phase oxidation of P25 mayake additional time for the •OH to reach the Or-II molecules inolution.

The performance of the photocatalyst in the photocatalyticisinfection of water supposes an aggressive test of their photo-xidation activity, since E. coli is a 1–3 �m living cell withrotection mechanisms to oxidative stress and 3.5 million macro-olecules on its cell wall. The survival bacteria after 40 min of illu-ination for TiO2 and Fe-modified TiO2 photocatalyst are shown

n Fig. 5. Blank experiments named as Dark and SODIS indicateo bactericidal activity of the system under darkness and a smallisinfection activity of simulated solar light respectively during0 min of irradiation. The performance of the photocatalysts inhe photocatalytic disinfection is comparable to their performancen the Or-II photocatalytic degradation, except for the activity ofT. Unlike the Or-II degradation, the E. coli inactivation by HT is

ower than that of P25. These results neglect the assumption thatT sample has a similar oxidative power than the one of the P25.evertheless, when dealing with enormous complex systems like. coli, compared to 1 nm Or-II molecule, interactions betweenhe target and the photocatalyst become highly important. Gumyt al. [4] found an important relation of the surface charge of thehotocatalyst and the cell wall charge. E. coli cells have positivend negative charged macromolecules on their surface withredominant negative groups. This implies that a more positiveharged surface is preferred to diminish in order to increase thelectrostatic attraction between the photocatalyst’s surface andacteria to be efficiently attacked and inactivated.

As proposed before, the oxidation activity of hydrothermallyynthesized photocatalysts (HT and Fe-HT) seems to be promotedy surface oxidation due to the h+. Since Kormali et al. [28] haveroposed that the photo-oxidation mode of the photocatalystepends on the nature of the substrate, and as it is well-known for25, it is possible that photocatalytic oxidation for the inactivationf E. coli in water is mainly caused by the attack of free •OHadicals to the cell wall in the homogenous phase. This assumptionn mechanisms may explain the difference in photocatalytic

isinfection activities for HT and P25 particles. It is possible tossume that HT particles have a high surface interaction withr-II to promote surface h+ oxidation mode, but there is no such

nteraction between HT surface and E. coli to promote a similarOH radical mode oxidation as P25 do.

Today 157 (2010) 119–124 123

Moreover, Fe enhances the photoactivity of the HT sampletowards the photocatalytic disinfection of water possibly due tothe new interaction bacteria-surface promoted by the presence ofFe. The presence of Fe is not toxic to E. coli cells since this element isa macronutrient for microorganisms; thus, the bactericidal activityof the Fe-HT sample is mainly due to photocatalytic reactions. Blankexperiments with Fe-HT showed final concentrations of 8.3 × 106

and 8.7 × 106 CFU/mL after 40 min of stirring under darkness, whichcorroborate the latter assumption.

The photocatalytic disinfection and photocatalytic detoxi-fication activities seem to be related to the crystalline structure,the optical properties, the chemical state of doping metal, and theinteractions between the surface of the photocatalyst and the tar-get. In this work we proposed a correlation of these characteristicsby comparing the photocatalytic activity of two different Fe–TiO2samples synthesized by well-known liquid synthesis methods:sol–gel and hydrothermal synthesis. This correlation allows us toassume that the synthesis route determines the activity of Fe dopedTiO2 since the chosen synthesis may promote the stabilizationof Fe in the TiO2 matrix enhancing the photoactivity. Literaturehas reported that iron doping in TiO2 increases photoactivity inthe visible region due to the intra band gap states promoted byFe3+ [29], but contrarily, Fe3+ may decrease photoactivity [29,30]under UV-light irradiation. These differences in activities mayconduct to lower activities under solar light irradiation where bothranges, UV and visible, are used. In this work, we present a feasiblesynthesis route to obtain Fe doped TiO2 particles with enhancedphotoactivity towards the degradation of the chosen targets undersolar light irradiation.

4. Conclusions

Crystalline TiO2 and Fe-modified TiO2 particles were synthe-sized by hydrolysis of titanium butoxide in the presence of aFe inorganic salt by the sol–gel and the hydrothermal synthe-sis. Hydrothermal synthesis leads to TiO2 particles with higherphotocatalytic oxidation activity towards the degradation of orangeII than TiO2 sol–gel synthesized particles and similar to the onesof P25. Results of characterization techniques indicate that thereis no direct correlation of the photoactivity with the crystallinityof anatase phase but to the presence of amorphous and the co-existence of anatase with one of two phases, rutile and brookite. Fewas effectively inserted in the TiO2 matrix by the chosen synthesisroutes. Diffusive reflectance analysis shows an increase in visiblelight absorption capacity due to the presence of Fe. However, thesol–gel process promotes Fe as a recombination center of the photo-generated e−–h+ reducing photoactivity. Contrarily, hydrothermalsynthesis leads to Fe–TiO2 particles with enhanced photocatalyticactivity, possibly due to the stabilization of Fe into the structureat low temperatures.

Additionally, the difference in performance of the photo-catalysts when degrading an organic molecule and, on the otherhand, inactivating bacteria suggests a difference in mechanismsof photo-oxidation by HT and P25. From analysis of the results,it is possible to assume a surface-hole-oxidation mode for hydro-thermally synthesized TiO2 particles and a homogeneous phaseoxidation mode promoted by •OH radicals produced in the TiO2P25 surface.

Acknowledgments

This work was financially supported by COLCIENCIAS and SENA(Project code. 1102341-19419). Camilo Castro also thanks thenamed Colombian government entities and UIS for the financialsupport for his PhD study.

1 alysis

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

24 C.A. Castro-López et al. / Cat

eferences

[1] J.M. Herrman, Heterogeneous photocatalysis: fundamentals and applicationsto the removal of various types of aqueous pollutants, Catal. Today 53 (1999)115–129.

[2] O. Carp, C.L. Huisman, A. Reller, Photoinduced reactivity of titanium dioxide,Prog. Solid State Chem. 32 (2004) 33–177.

[3] J. Blanco, P. Fernández, S. Malato, Solar photocatalytic detoxification and dis-infection of water: recent overview, J. Sol. Energy Eng. 129 (2007) 4–15.

[4] D. Gumy, C. Morais, P. Bowen, C. Pulgarín, S. Giraldo, R. Hajdu, J. Kiwi, Catalyticactivity of commercial TiO2 powders for the abatement of the bacteria (E. coli)under solar simulated light: influence of the isoelectric point, Appl. Catal. B:Environ. 63 (2006) 76–84.

[5] M. Mahalaksmi, S. Vishnu, B. Arabindoo, M. Palanichamy, V. Murugesan, Photo-catalytic degradation of aqueous propoxur solution using TiO2 and H� zeolitesupported TiO2, J. Hazard. Mater. 161 (2009) 336–343.

[6] W.Y. Choi, A. Termin, M.R. Hoffmann, The role of metal ion dopants inquantum-sized TiO2: correlation between photoreactivity and charge carrierrecombination dynamics, J. Phys. Chem. 98 (1994) 13669–13679.

[7] M.I. Litter, J.A. Navío, Photocatalytic properties of iron doped-titania semicon-ductors, J. Photochem. Photobiol. A: Chem. 98 (1996) 171–181.

[8] T.K. Ghorai, S.K. Biswas, P. Pramanik, Photo-oxidation of diferent organic dyes(RB, MO, TB, and BG) using Fe(III)-doped TiO2 nano photocatalyst prepared bynovel chemical method, Appl. Surf. Sci. 254 (2008) 7498–7504.

[9] B. Xin, Z. Ren, P. Wang, J. Liu, L. Jing, H. Fu, Study on the mechanisms of photoin-duced carriers separation and recombination for the Fe3+–TiO2 photocatalysts,Appl. Surf. Sci. 253 (2007) 4390–4395.

10] M. Asiltürk, F. Sayilkan, E. Arpac, Effect of Fe3+ ion doping to TiO2 on the pho-tocatalytic degradation of Malachite green dye under UV and vis irradiation, J.Photochem. Photobiol. A: Chem. 203 (2009) 64–71.

11] M.A. Khan, S.I. Woo, O. Yang, Hydrothermally stabilized Fe(III) doped titaniaactive under visible light for water splitting reaction, Int. J. Hydrogen Energy33 (2008) 5345–5351.

12] S. Liu, Y. Chen, Enhanced photocatalytic activity of TiO2 powders doped by Feunevenly, Catal. Commun. 10 (2009) 894–899.

13] C.C. Trapalis, P. Keivanidis, G. Kordas, M. Zaharescu, M. Crisan, A. Szatnayi, M.Gartner, TiO2 (Fe3+) nanostructured thin films with antibacterial properties,Thin Solid Films 433 (2003) 186–190.

14] C.Y. Wang, C. Bottcher, D.W. Bahneman, J.K. Dohrman, A comparative study ofnanometer size Fe(III)-doped TiO2: photocatalysts: synthesis, characterization

and activity, J. Mater. Chem. 13 (2003) 2322–2325.

15] Y. Xie, C. Yuan, X. Li, Photosensitized and photocatalyzed degradation of azo dyeusing Lnn+–TiO2 sol in aqueous solution under visible light irradiation, Mater.Sci. Eng. B 117 (2005) 325–333.

16] C. Adan, A. Martinez, S. Malato, A. Bahamonde, New insights on solarphotocatalytic degradation of phenol over Fe–TiO2 catalysts: photo-

[

[

Today 157 (2010) 119–124

complex mechanism of iron lixiviates, Appl. Catal. B: Environ. 93 (2009)96–105.

17] D.C. Hurum, A.G. Agrios, K.A. Gray, T. Rajh, M.C. Thurnauer, Explaining theenhanced photocatalytic activity of Degussa P-25 mixed phase oxide TiO2 usingEPR, J. Phys. Chem. 107 (2003) 4545–4549.

18] S. Ardizzone, C. Bianchi, G. Cappelletti, S. Gialanella, C. Pirola, V. Ragaini, Tai-lored anatase/brookite nanocrsytalline TiO2. The optimal particle features forliquid and gas-phase photocatalytic reactions, J. Phys. Chem. C 111 (2007)13222–13231.

19] A. Paola, G. Cufalo, M. Addamo, M. Bellardita, R. Campostrini, M. Ischia, R. Cec-cato, L. Palmisano, Photocatalytic activity of nanocrystalline TiO2 (brookite,rutile, and brookite-based) powders prepared by thermohydrolysis of TiCl4in aqueus chloride solutions, Colloids Surf. A: Physicochem. Eng. Aspects 317(2008) 366–376.

20] P. Pitter, J. Chudoba, Biodegradability of Organic Substances in the AquaticEnvironment, CRC Press, Boca Raton, 1990.

21] G.P. Anipsitakis, T.P. Tufano, D. Dionysiou, Chemical and microbial decontami-nation of pool water using activated potassium peroxymonosulfate, Water Res.42 (2008) 2899–2910.

22] C. Karunakaran, P. Anilkumar, Semiconductor-catalyzed solar photo-oxidationof idodide ion, J. Mol. Catal. A: Chem. 265 (2007) 153–158.

23] N. Todorova, T. Giannakopoulou, G. Romanos, T. Vaimakis, J. Yu, C. Trapalis,Preparation of fluorine-doped TiO2 photocatalysts with controlled crystallinestructure, Int. J. Photoenergy 2008 (2008) 1–9.

24] K.Y. Jung, S.B. Park, Anatase-hase titania: preparation by embedding silicaand photocatalytic activity forthe decomposition of trichloroethylene, J. Pho-tochem. Photobiol. A: Chem. 127 (1999) 117–122.

25] I.K. Konstantinou, T.A. Albanis, TiO2-assisted photocatalytic degradation of azodyes in aqueous solution: kinetic and mechanistic investigations. A review,Appl. Catal. B: Environ. 49 (2004) 1–14.

26] T. Tong, J. Zhang, B. Tian, F. Chen, D. He, Preparation of Fe3+-doped TiO2 catalystby controlled hydrolisis of titanium alcoxide and study on their photocatalyticactivity for methyl orange degradation, J. Hazard. Mater. 155 (2008) 572–579.

27] C. Guillard, H. Lachheb, A. Houas, M. Ksibi, E. Elaloui, J.M. Herrman, Influenceof chemical structure of dyes, of pH and inorganic salts on their photocatalyticdegradation by TiO2 comparison of the efficiency of powder and supportedTiO2, J. Photochem. Photobiol. A: Chem. 158 (2003) 27–36.

28] P. Kormali, T. Triantis, D. Dimotikali, A. Hiskia, E. Papaconstantinou, On thePhoto-oxidative behavior of TiO2 and PW12O40

3−: OH radicals versus holes,Appl. Catal. B: Environ. 68 (2006) 139–146.

29] Y. Wu, J. Zhang, L. Xiao, F. Chen, Preparation and characterization of TiO2 pho-tocatalyst by Fe3+ doping together with Au deposition of organic pollutants,Appl. Catal. B: Environ. 88 (2008) 525–532.

30] Z. Li, W. Shen, W. He, X. Zu, Effect of Fe-doped TiO2 nanoparticle derivedfrom modified hydrothermal process on the photocatalytic degradation per-formance on methylene blue, J. Hazard. Mater. 155 (2008) 590–594.